Petrologic Analysis of Green-Black Impact Melt Rock with a History Of

50th Lunar and Planetary Science Conference 2019 (LPI Contrib. No. 2132) 1718.pdf

PETROLOGIC ANALYSIS OF GREEN–BLACK IMPACT MELT ROCK WITH A HISTORY OF HYDROTHERMAL ALTERATION AT CHICXULUB. Stephen J. Slivicki1,2, Martin Schmieder1, David A. Kring1 and the IODP–ICDP Expedition 364 Science Party, 1Lunar and Planetary Institute, 3600 Bay Area Blvd., Houston, TX 77058 USA; 2Dept. of Plant & Earth Science, University of Wisconsin River Falls, River Falls, WI 54022 USA ([email protected]).

Introduction: With a diameter of ~180 km, the were analyzed using a Leica optical microscope at the end-Cretaceous Chicxulub crater is one of the largest LPI and a JEOL 5910 LV scanning electron impact structures on Earth [1,2]. The impact occurred microscope (SEM) and CAMECA SX 100 electron on a carbonate-evaporite platform on top of a complex microprobe at the NASA Johnson Space Center. crustal basement and generated large volumes of Petrography and Geochemistry: The bulk of the impact melt [1–3]. At Chicxulub, impact melt is of black domain is a dark, aphanitic impact melt rock with andesitic composition and concentrated in a central target rock clasts up to several cm in size. The micro- melt pool inside the peak ring, as well as in an annular crystalline melt groundmass contains mainly feldspar, trough surrounding the peak ring [1,4–6]. The crater pyroxene, and opaque minerals. Groundmass feldspar hosted a spatially extensive post-impact hydrothermal is An0–50Ab1–70Or3–99 (n=32). Pyroxene is augite, system with an estimated lifetime of ~1.5–2.3 Myr that Wo38–50En22–46Fs28–11 (n=60; FeO/MnO [wt%]=18–54). pervasively altered Chicxulub’s impactites [3]. Samples (1) and (4) have a pyroxene-rich melt Evidence for this is found in borehole samples from the groundmass, while the groundmass of samples (2) and crater, such as Yucatán-6 (Y-6) and Yaxcopoil-1 (3) contains far more feldspar; in those samples, (Yax-1) [6-10]. More recently, IODP–ICDP Expedi- pyroxene is mainly found in coronas surrounding tion 364 recovered drill core M0077A from the north- quartz clasts (as in [6]). The black melt domain con- western outer flank of the peak ring, which sampled tains abundant mineral and rock clasts, most of which basement lithologies and impactites [2]. Of particular are siliceous in nature. Many siliceous clasts exhibit interest for this study is a zone of black–green impact shock textures (e.g., PDFs in quartz, ballen quartz, and melt rock within that core between 721 and 747 meters toasted quartz [11]). Vesicles are abundant in the black below sea floor (mbsf) that shows two distinct domains domain and commonly contain fillings of secondary (Fig. 1) [2]. Our goal is to resolve whether those two smectite-group clay minerals (likely saponite) and lo- domains represent (1) intermingled types of (now cally K-feldspar (adularia), pyrite, magnetite, and spar- altered) impact melt of different original composition; ry calcite. Vesicles and cracks, e.g., in sample (1), are (2) crystallized black impact melt juxtaposed with a commonly surrounded by halos of K-feldspar. domain of localized hydrothermal alteration; or Samples (5) and (6) represent the black melt rock perhaps a combination thereof. interspersed with a conspicuous green, pseudofluidal (schlieren-like) alteration domain that consists of ~60% sheet silicate, ~25% calcite (locally intergrown with microfibrous zeolite), ≤15% garnet, and minor opaque minerals. As revealed by SEM imaging, the sheet silicate has a fibrous to flaky microtexture and is rich in Mg, Fe, Al, and Ca, suggesting it is likely saponite, a smectite-group clay mineral [7,12-15]. Electron microprobe results suggest the green domain, vesicle- filling clays in the black domain, and clay-altered im- Fig. 1: Core M0077A segment 88–4 (726 mbsf) from IODP– ICDP Expedition 364 displaying both green and black domains pact glass from the overlying suevite unit [2] are simi- of a hydrothermally-altered impact melt unit. Scale is in cm. lar in composition (Fig. 2). Garnet in the green domain

Samples and Analytical Methods: Six polished occurs in the form of euhedral, yellowish-translucent thin-section samples of the black–green melt rock in- crystals, with andradite cores (And62–72Gro35–28Uvt0–3) terval in Expedition 364 core M0077A were examined that have ≤12 wt% TiO2 and grossular-rich rims in this study: (1) 85-1-26-28 (717 mbsf); (2) 88-3-48.5- (Gro72–82And28–18), and seems to be a pervasive hydro- 50 (725 mbsf); (3) 89-3-13.5-15 (728 mbsf); (4) 90-3- thermal mineral within the green domain. The green 70.5-72 (732 mbsf); (5) 89-3-39-43 (729 mbsf); and domain entrains clasts or relics of the black melt rock (6) 92-2-91.5-93 (737 mbsf). Samples (1)–(4) repre- and, more rarely, clasts of microcrystalline-spherulitic sent the ‘black’ melt domain while samples (5) and (6) silica and limestone. It is, overall, poorer in clasts than are from the ‘green’ alteration domain and the bounda- the black domain and the boundary between the two ry between the two domains (Fig. 1). Thin sections domains commonly appears to be relatively sharp. 50th Lunar and Planetary Science Conference 2019 (LPI Contrib. No. 2132) 1718.pdf

Discussion: Impactites on top of the Chicxulub material in melt-bearing breccias elsewhere in the core peak ring record a complex history of impact melt for- (Fig. 2). Such zones of (near-)glassy schlieren within mation and hydrothermal alteration [8]. The composi- the black melt, potentially with a geochemical compo- tion of groundmass pyroxene and feldspar in the black sition slightly different from that of the surrounding melt domain seems to be homogeneous across samples melt, may have been preferentially altered by hydro- from core M0077A and is similar to that of pyroxene in thermal fluids. The zones may have been slightly impact melt-bearing lithologies from other drill core enriched in Ca, producing andradite garnet upon altera- sites at Chicxulub [6,7,9]. This suggests the composition, but not wholly carbonate melts. The clay-rich, tion of impact melt within the crater was relatively uni- silicate-dominated green zones have a relatively low form, while cooling and alteration in different parts of abundance of calcite and lack typical liquid silicate– the crater may have been variable. Pervasive hydro- carbonate immiscibility textures (e.g., [20]) and, thus, thermal alteration in the black melt domain is evi- are inconsistent with a carbonate melt-dominated denced by K-metasomatism along cracks and vesicles origin. [8] and secondary vesicle fillings. The green domain may, alternatively, represent a The green domain, in contrast, almost entirely con- purely hydrothermal filling of large, elongated vesicles, sists of secondary minerals that formed in a proposed and/or open fractures within the black melt rock, simi- post-impact crystallization sequence: garnet → smec- lar to a model for saponite crystallization proposed tite → calcite [8]. While hydrothermal garnet is rare in previously for impactites in the Yax-1 core [17]. The other Chicxulub core samples [16,17], it is locally a filling of cavities and fractures would have occurred rock-forming mineral in the green-altered melt zones of while hydrothermal fluids circulated through a permea- core M0077A. The presence of garnet suggests mini- ble unit [17]. However, the presence of lithic clasts and mum temperatures of ~280 to 300°C in the crater- fragments of black melt in the green domain is incom- hosted hydrothermal system [8,16–18], whereas patible with that being solely the origin of those zones. saponite probably formed at ~100 to 150°C [19]. Conclusions: The intermingled green and black domains within the impact melt unit of M0077A probably represent two slightly different textural and/or compositional melts that subsequently reacted to hydrothermal fluids in different ways. The green domain was probably finer-grained, if not glassy, compared with the black domain, and, thus, more susceptible to alteration. The alteration produced a saponite matrix similar to altered glasses elsewhere in the core, aug- mented with a spectacular array of grossular-rimmed andradite garnet. References: [1] Kring D. A. (2005) Chemie d. Erde, 65, 1–46. [2] Morgan J. V. et al. (2016) Science, 354, 878– 872. [3] Abramov O. and Kring D. A. (2007) MAPS, 42, 93– Fig. 2: Composition of mafic sheet silicates in black-green 112. [4] Sharpton V. L. et al. (1996) GSA Spec. Pap., 307, impact melt domain (green diamonds), indicating saponitic 55–74. [5] Schuraytz B. C. et al. (1994) Geology, 22, 868– compositions (Sap; gray field). Smectite-altered impact glass in 872. [6] Kring D. A. and Boynton W. V. (1992) Nature, 358, the overlying suevite unit (triangles) has a similar composition. 141–144. [7] Zürcher L. and Kring D. A. (2004) MAPS, 39, Vermiculite (Vrm) and montmorillonite (Mnt) are shown for comparison [7,11-14]. The inset SEM image shows an example 1199–1221. [8] Kring D. A. et al. (2017) LPS XLVIII, abstr. of green clay alteration in core interval 90–2 (depth 731 mbsf). #1212. [9] Ames D. E. et al. (2004) MAPS, 39, 1145–1167. [10] Hecht L. et al. (2004) MAPS, 39, 1169–1186. Due to pervasive hydrothermal alteration, the [11] French B. M. (1998) Traces of Catastrophe, LPI source ‘protolith’ of the green domain is not immedi- Contrib. 954, 120 pp. [12] Handbook of Mineralogy, online. ately obvious. The green domain may represent an [13] Deer W. A., Howie R. A. and Zussman J. (1992, Eds.) nd alteration product of a finer-grained or glassy silicate- An Introduction to the Rock-Forming Minerals, 2 ed., rich melt that was once intermingled with the surviving Longman, 696 pp. [14] Treiman A. H. et al. (2014) Am. Mineral., 99, 2234–2250. [15] Sætre C. et al. (2018) MAPS black crystalline impact melt. This type of subtle (in press), doi: 10.1111/maps.13214. [16] Newsom H. E. et bimodal textural association of melt occurs in rapidly al. (2010) LPS XLI, abstr. #1751. [17] Nelson M. J. et al. quenched volcanic melts, which may be a suitable (2012) GCA, 86, 1–20. [18] McCarville P. and Crossey L. J. thermal and crystallization analogue for these impact (1996) GSA Spec. Pap., 302, 347–376. [19] Osinski G. R. melts. This interpretation is consistent with similar (2005) Geofluids, 5, 202–220. [20] Osinski G. R. and Spray saponite filling vesicles in the black melt domain and J. G. (2001) EPSL, 194, 17–29. with similarly-altered, saponite clasts of once-glassy